Unverified50% confidenceFactExact time
Vectorizing a tensor destroys the multidimensional structure of the data and leads to an explosion in the number of parameters; if a p1×p2×…×pK tensor is directly flattened into a vector, its dimension is the product of the dimensions of all modes, and the number of parameters grows exponentially with the number of modes.
1
Sources
50%
Confidence
Long-term
Relevance
10/27/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIf each layer of delegation appends a complete assertion, token size grows linearly with the delegation depth, causing the assertion bloat problem.73% similarUnverifiedFlattening a matrix and feeding it to an MLP results in a non-fixed input dimension due to variable row counts, and breaks permutation invariance.72% similarUnverified数值微分计算量与参数数量成正比，符号微分会产生表达式膨胀问题，二者在大规模网络中都无法扩展，反向传播是解决方案70% similarUnverifiedChunk size is a hyperparameter that requires trade-offs: chunks that are too small result in incomplete semantics for a single segment, while chunks that are too large introduce more noise, occupy a longer context window, and increase inference costs.70% similarUnverifiedThe computational complexity of diffusion models in the temporal dimension grows approximately quadratically with duration, and memory and compute consumption rise sharply.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1102404API
curl https://kongchang.com/api/v1/knowledge/claims/1102404MCP
get_claim(id=1102404)